Industrial polymerization process quality prediction based on CPSO-LSTM-RNN

Xuesong Wang, Cuimei Bo, Jun Li

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Due to the industrial polymer process is a complex and nonlinear process with the characteristic of multi-variables, hysteresis, large inertia and strong coupling, the main production targets are difficult to measure accurately, and there are fluctuations in data information during the production, so many production data need to be analyzed and processed. Therefore, this paper proposes the CPSO-LSTM-RNN algorithm to predict the yield of industrial polymer process. Firstly, the LSTM-RNN model is established and the model is trained with the data of the production process. Then, the CPSO algorithm is used to obtain the optimal hyperparameters of the model. Finally, the validity of the model is verified by a set of industrial data of the polymerization process.

Original languageEnglish
Title of host publicationECITech 2022 - 2022 International Conference on Electrical, Control and Information Technology
EditorsMarco C. Campi, Ning Wang
PublisherVDE VERLAG GMBH
Pages808-811
Number of pages4
ISBN (Electronic)9783800759170
StatePublished - 2022
Event2022 International Conference on Electrical, Control and Information Technology, ECITech 2022 - Virtual, Online
Duration: 25 Mar 202227 Mar 2022

Publication series

NameECITech 2022 - 2022 International Conference on Electrical, Control and Information Technology

Conference

Conference2022 International Conference on Electrical, Control and Information Technology, ECITech 2022
CityVirtual, Online
Period25/03/2227/03/22

Fingerprint

Dive into the research topics of 'Industrial polymerization process quality prediction based on CPSO-LSTM-RNN'. Together they form a unique fingerprint.

Cite this